Skip to main content
Skip to footer

Unnati Kohli

Data is the currency of the digital economy. With an increasing volume of data flowing in from multiple sources, enterprises are banking on artificial intelligence (AI), machine learning (ML), and natural language processing (NLP) technologies to drive data-driven decision-making.

The amount of data emerging from next-generation telecom networks such as 5G, edge, ORAN, and so on, is unprecedented. Telecommunication networks–from wireless to fixed line–are fed with system-generated structured data from a variety of data sources. Although structured data is believed to provide better insights, utilizing unstructured data is equally crucial to gather actionable information.

One of the best sources of unstructured data is the helpdesk ticketing system. The unanalyzed service help-desk data in the database of a telecom company includes details such as customer issues, time taken to close a ticket, and the resolution provided. When this ticket data is integrated with the Networks Data Analytics Function (NWDAF) in 5G defined by the 3rd Generation Partnership Project (3GPP), it can improve the customer experience immensely. The Networks Data Analytics Function is a repository of extensive data collected from numerous end devices such as mobiles, network elements, drones, IoT applications, and so on. It is primarily used to gather structured data from various network functions and provide analytical insights.

With in-built machine learning capabilities, NWDAF is also a vital component in 5G network analytics that enables network service providers to gather the help desk ticket data seamlessly in one place, implement their own machine learning algorithms, and draw relevant insights. This drives smart ticketing. Further, integrating machines with a network domain-guided knowledge database (KDB) and NLP processing algorithms can help draw critical insights from the ticket data. It has a significant impact on customer service as it helps predict downtime and categorize tickets for timely and focused resolution.

Let us look at a few use cases for handling the data gathered from NWDAF and service tickets:

  • Leveraging intelligent routing of tickets to the right team or department by categorizing the tickets into limited types such as hardware, software, and coverage; this ensures adherence to service level agreements (SLAs) and significantly decreases the average handling time (AHT)
  • Combining the tickets data with real-time social media data to gather insights on customer needs and behaviors; end users often mention the names of competitors as well, which can provide competitor insights to telecom companies
  • Evaluating and analyzing the pattern in tickets (if there is any), and identifying the association between the related network elements can enable companies to do a causal for the issues

There are multiple ways to efficiently deploy NWDAF with next-generation analytics. It is critical to opt for a distributed architecture for analytics as it allows real-time usage and decreases the overhead network bandwidth. It is essential to select modular analytics systems that allow customization of NWDAF and ensure that the vendors have NWDAF support to receive analytics services. Additionally, storing smart data instead of unprocessed big data, choosing the analytics solutions in adherence to the five nines SLAs, and opting for cloud solutions to reduce expenses are important. It is also vital to ensure that the analytics use cases are aligned to the constantly changing nature of the telecom business.

The integration of 5G NWDAF with NLP in the telecom industry has numerous potential business benefits. By leveraging the evolving technologies and analytics, companies can improve network performance, drive closed-loop network automation, deliver innovative services, and maximize profit generation. They can take corrective actions by predicting deterioration in network and preventing performance issues in addition to gaining effective returns on network capacity investments. Using the potential of network technologies, telecom companies can easily adapt to the shifting network conditions and balance network loads while assuring quality of experience.

About the author

Unnati Kohli
Unnati Kohli is a data analyst with TCS' Communications, Media and Information Services (CMI) business unit. She works with TCS’ customers to design solutions using artificial intelligence, machine learning and analytics technologies to improve their network operations strategy and resolve business challenges.
Contact Contact